I installed CUDA by ]add CUDA and tried to test it by test CUDA.
However, I encountered the below error message.
CUDA initialisation error
ERROR: LoadError: CUDA.jl did not successfully initialize, and is not usable.
If you did not see any other error message, try again in a new session
with the JULIA_DEBUG environment variable set to 'CUDA'.
And this message is popped up every time I try to use any functionality of CUDA, for example,
cu error
julia> using CUDA
julia> cu(rand(5))
ERROR: CUDA.jl did not successfully initialize, and is not usable.
If you did not see any other error message, try again in a new session
with the JULIA_DEBUG environment variable set to 'CUDA'.
Stacktrace:
[1] error(::String) at ./error.jl:33
[2] libcuda() at /home/jinrae/.julia/packages/CUDA/YeS8q/src/initialization.jl:51
[3] (::CUDA.var"#690#cache_fptr!#9")() at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/utils/call.jl:31
[4] macro expansion at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/utils/call.jl:39 [inlined]
[5] macro expansion at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/cudadrv/libcuda.jl:29 [inlined]
[6] macro expansion at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/cudadrv/error.jl:102 [inlined]
[7] cuDeviceGet(::Base.RefValue{Int32}, ::Int64) at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/utils/call.jl:93
[8] CuDevice(::Int64) at /home/jinrae/.julia/packages/CUDA/YeS8q/lib/cudadrv/devices.jl:25
[9] initialize_thread(::Int64) at /home/jinrae/.julia/packages/CUDA/YeS8q/src/state.jl:121
[10] prepare_cuda_call() at /home/jinrae/.julia/packages/CUDA/YeS8q/src/state.jl:80
[11] device at /home/jinrae/.julia/packages/CUDA/YeS8q/src/state.jl:227 [inlined]
[12] alloc at /home/jinrae/.julia/packages/CUDA/YeS8q/src/pool.jl:293 [inlined]
[13] CuArray{Float32,1}(::UndefInitializer, ::Tuple{Int64}) at /home/jinrae/.julia/packages/CUDA/YeS8q/src/array.jl:20
[14] CuArray at /home/jinrae/.julia/packages/CUDA/YeS8q/src/array.jl:76 [inlined]
[15] similar at ./abstractarray.jl:675 [inlined]
[16] convert(::Type{CuArray{Float32,N} where N}, ::Array{Float64,1}) at /home/jinrae/.julia/packages/GPUArrays/jhRU7/src/host/construction.jl:82
[17] adapt_storage at /home/jinrae/.julia/packages/CUDA/YeS8q/src/array.jl:330 [inlined]
[18] adapt_structure at /home/jinrae/.julia/packages/Adapt/8kQMV/src/Adapt.jl:42 [inlined]
[19] adapt at /home/jinrae/.julia/packages/Adapt/8kQMV/src/Adapt.jl:40 [inlined]
[20] cu(::Array{Float64,1}) at /home/jinrae/.julia/packages/CUDA/YeS8q/src/array.jl:342
[21] top-level scope at REPL[7]:1
Note: this is the result of nvidia-smi. It seems not use GPU at all when I tried to train a neural network by Flux.
What should I do?
nvidia-smi
➜ GliderPathPlanning git:(using-gpu) ✗ nvidia-smi
Fri Jan 1 00:28:32 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 455.45.01 Driver Version: 455.45.01 CUDA Version: 11.1 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 GeForce RTX 207... Off | 00000000:01:00.0 Off | N/A |
| N/A 54C P8 10W / N/A | 1500MiB / 7982MiB | 34% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1092 G /usr/lib/xorg/Xorg 134MiB |
| 0 N/A N/A 1852 G /usr/lib/xorg/Xorg 840MiB |
| 0 N/A N/A 2040 G /usr/bin/gnome-shell 230MiB |
| 0 N/A N/A 2306 G ...mviewer/tv_bin/TeamViewer 2MiB |
| 0 N/A N/A 673753 G ...AAAAAAAA== --shared-files 100MiB |
| 0 N/A N/A 673779 G ...AAAAAAAA== --shared-files 176MiB |
+-----------------------------------------------------------------------------+
CUDA.has_cuda()
julia> CUDA.has_cuda()
false
